Blood Flow Velocity of Middle Cerebral Artery during Prolonged Anesthesia with Halothane, Isoflurane, and Sevoflurane in Humans

The study protocol was approved by the Ethical Committee for Human Study of the Yamaguchi Rosai Hospital, and informed consent was obtained from each patient. Twenty-four (American Society of Anesthesiologist' (ASA) physical status I or II patients (15 men, 9 women) were randomly assigned to three groups (8 for each) receiving either halothane, isoflurane, or sevoflurane during elective surgery. Ages of the patients ranged from 20 to 73 yr. Surgical procedures included orthopedic and abdominal surgery, which lasted more than 3 h. Atropine sulfate, 0.5 mg, and midazolam, 3-5 mg, were given intramuscularly 30 min before induction. Anesthesia was induced with the selected volatile anesthetic in an air-oxygen mixture adjusted to obtain FiO²of 0.35, and inspired concentration of volatile anesthetic was increased to 3-4% over 3-4 min. Tracheal intubation was facilitated with intravenous administration of 8-mg vecuronium bromide. After intubation, end-tidal concentration of the selected volatile anesthetic was adjusted to an age-appropriate level of 0.5 MAC, [16-19]then increased to 1.0 MAC, 1.5 MAC before surgery, and maintained at 1.5 MAC for the period of surgical procedure. After the end of surgery, the concentration of volatile anesthetics was stepwise decreased (1.5, 1.0, and 0.5 MAC).

Patients' lungs were mechanically ventilated to maintain normocapnia, and FiO²was kept at 0.35. The end-tidal concentrations of carbon dioxide (ETCO²) and volatile anesthetic were continuously monitored with a calibrated infrared gas analyzer (Capnomac Ultima, Datex, Helsinki, Finland). Apart from vecuronium, no other drugs were administered during the surgical procedure except phenylephrine, which was used to maintain mean arterial blood pressure (MABP, see below). The nasopharyngeal temperature was monitored by a calibrated thermistor probe and was kept at 35.5-37.0 [degree sign] Celsius using a cooling-warming water mattress. Bilateral unipolar (earlobe as a reference electrode), frontal, and parietal electroencephalographs (EEGs) were monitored and recorded continuously (Neuropack 8, Nihon Kohden, Tokyo, Japan). The electrocardiograph (ECG) also was monitored. A 22-gauge Teflon indwelling catheter was placed in the radial artery for blood sampling and pressure measurement. The arterial pressure was measured by strain gauge transducers with the zero point at the mastoid process and was recorded on a polygraph (Lifescope 14, Nihon Kohden, Tokyo). The MABP was maintained above 70 mmHg with a continuous infusion of phenylephrine (0.1-1.0 micro gram [center dot] kg sup -1 [center dot] min sup -1), if necessary. Arterial blood samples were obtained every 60 min and analyzed for oxygen tension (PO²) and carbon dioxide tension (PCO²) with a blood gas analyzer (ABL505, Radiometer, Copenhagen, Denmark) at 37.0 [degree sign] Celsius. Hemoglobin concentration was measured spectrophotometrically (OSM3, Radiometer).

A 2-MHz pulsed Doppler ultrasound device (Transscan, EME, Uberlingen, Germany) was used for transcranial measurements of blood flow velocity of the right MCA. TCD ultrasound signals were identified at a depth of 42-60 mm. Meticulous care was taken to ensure a constant position of the ultrasound probe by use of a suitable holder attached to the patient's head (IMP-2 monitoring probe holder, EME). Measurement of flow velocity was started shortly after induction of anesthesia. Flow velocity and blood pressure were displayed simultaneously on a video screen and continuously recorded on a microcomputer by use of the long-term monitoring option of the Doppler device. V^mcawas derived from on-line integration of the recordings of envelope curves of maximal intravascular velocity. During induction or emergence of anesthesia, equilibration time at each end-tidal concentration is 5-10 min. During operation at 1.5 MAC, V^mcarepresents an average value of those recorded for approximately 1 min at each time point (every 15 min). When EEG shows burst suppression pattern, we calculated averaged value from both V^mca, values during burst period and suppression period.

Data are expressed as mean +/- SD. Between-group comparisons of demographic data were made by one-way analysis of variance (ANOVA). Gender distribution between groups was compared by chi-square analysis. For the statistical analysis on physiologic variables and V^mca, the data were separated into three parts: the initial dose-response data (induction period, 0.5, 1.0, 1.5 MAC), the time course data at 1.5 MAC, and the final dose-response data (emergence period, 1.5, 1.0, 0.5 MAC). Two-way ANOVA for repeated measures was applied on each three part. Bonferroni's post hoc test was applied for between-group comparisons as indicated. Statistical significance was assumed when P < 0.05.

A summary of demographic data is shown in Table 1. There were no significant differences among three groups. Figure 1showed MABP and heart rate during the study. There were no significant differences among three groups except MABP during induction: MABP in the sevoflurane group was significantly (P < 0.05) lower than those in the halothane and isoflurane groups during induction. The total number of patients given phenylephrine was eight, three, and four in the halothane, isoflurane, and sevoflurane groups, respectively. Hemoglobin concentration (9.4-14.6 g/dl), PaO²(99-250 mmHg), PaCO²(37-45 mmHg), ETCO²(35-48 mmHg), and nasopharyngeal temperature (35.7-37.5 [degree sign] Celsius) were stable in each patient. No statistical significant differences were observed in these data among three groups.

(Figure 2) shows the time course of the changes of V^mca. V^mcaat 0.5 MAC of halothane, isoflurane, and sevoflurane was 49 +/- 19, 57 +/- 8, and 48 +/- 13 cm/s, respectively, with no significant difference among the anesthetics. Halothane significantly (P < 0.01) increased V^mcain a dose-dependent manner (0.5, 1.0, 1.5 MAC) during induction, whereas other two anesthetics produced no dose-related increase. During 3 h anesthesia period at 1.5 MAC, ANOVA for repeated measures showed significance (P < 0.05) for the time trends for V^mcadata in all groups, but there was no decay in V^mcaover time (Figure 2). No significant differences of V^mcawere obtained among three groups at 1.5 MAC during operation. During emergence of anesthesia, V^mcasignificantly (P < 0.01) decreased in a dose-dependent manner (1.5, 1.0, 0.5 MAC) in the halothane group, whereas no significant dose-related decrease was observed in the isoflurane and sevoflurane groups.

At 1.5 MAC, EEG recordings showed frequent burst (sharp or spike wave, 50-200 micro V) and suppression in the isoflurane group, whereas in the halothane and sevoflurane groups, 12-15 Hz activities (50-90 micro V) and 9-13 Hz activities (70-100 micro V) were predominant, respectively. In the sevoflurane group, burst suppression was observed in a short time in 2 patients.

(Figure 3) depicts representative recordings of MABP, V^mca, and EEG in a patient anesthetized with 1.5 MAC isoflurane. During burst and suppression in EEG, the onset of a burst resulted in a substantial increase (approximately 5-30 cm/s) in V^mca, which was maintained for the duration of the burst in all patients.

This study demonstrated in humans that V^mcasignificantly (P < 0.01) increased during induction or decreased during emergence in a dose-dependent manner in halothane anesthesia, whereas isoflurane and sevoflurane produced no significant dose-dependent change. Changes of V^mcaduring 3-h period at 1.5 MAC were significant (P < 0.05) for the time trends for all anesthetics, but there was no decay in V^mcaover time. It was also shown that Vmca fluctuated during burst suppression in EEG at 1.5 MAC of isoflurane (and sevoflurane), V^mcabeing increased during burst activity and decreased during suppression.

Dose-related increase in V^mcaobserved in the halothane group is in good agreement with that reported by Thiel et al. [20]With isoflurane, Bisonnette [21]showed no difference of V^mcabetween 0.5 and 1.5 MAC, and Thiel [20]showed no difference between 2 MAC and awake values. There have been many studies describing that the increase in CBF in isoflurane anesthesia, if present at all, is smaller than the changes observed with halothane. [22]Although the studies for sevoflurane are limited in number, either decreased [23,24]or unchanged [25,26]CBF was reported in animals. Low CBF values were reported in patients with ischemic cerebrovascular disease who were anesthetized with 1.5% sevoflurane and 33% nitrous oxide. [27]Recent study in healthy patients showed that V^mcawas decreased under 1.2 MAC sevoflurane anesthesia when compared with the awake value. [28]Although we have no awake value of V^mca, considering these previously mentioned results, no dose-related increase in the current study may be the case for isoflurane and sevoflurane.

Many animal studies using dogs and goats revealed that cerebral hyperemia induced by isoflurane and halothane spontaneously decreased over time. [1-6]However, some reports [7,8]have failed to observe the gradual decrease in CBF over time. There is no clear explanation for the different results reported in animals, but it may be a result of the differences of species or the technique of CBF measurement. It has been our concern whether time-dependent decay of CBF takes place in humans. There has been limited human investigation on this matter, and reported results suggest that CBF does not decrease over time during anesthesia. [9,10]However, this lack of time decay was based on two determinations in the beginning and at the end of the study only (in 1- and 2-h, [9]and 3- and 6-h [10]periods after induction). In our previous study, [11]we found that the elevated CBF equivalent, calculated every 20 min, was preserved during prolonged anesthesia (3 h) with volatile anesthetics and suggested that the increase of CBF was maintained over time. However, this indirect method has limitations: real CBF cannot be obtained, and only the global ratio of CBF-metabolism for certain time period is obtained. In the present study, we incorporated TCD technique to measure V^mcacontinuously during prolonged anesthesia. Flow velocity is not equivalent to CBF, but it is well accepted that the changes in flow velocity reflect corresponding relative changes in CBF. [12,13]Because the flow velocity with isoflurane and sevoflurane did not increase in a dose-dependent manner, it is difficult to state that an "increase" in flow velocity is maintained during isoflurane and sevoflurane anesthesia. However, no decay in V^mcaover time in all groups in this study suggests that the gradual decrease in CBF over time do not take place in humans.

The statistical significance of the changes in V^mcafor the time trends during 3-h period at 1.5 MAC was interpreted as a reflection of the fluctuation of V^mca, possibly in association with functional changes. The changes of EEG activities reflect the functional changes of the brain and may be associated with the change of CBF. During burst suppression pattern in EEG observed at 1.5 MAC isoflurane (and sevoflurane) anesthesia, the onset of a burst activity was associated with an increase of V^mca, which was maintained for the duration of the burst, and V^mcadecreased during the suppression period (Figure 3). This observation coincides with that by Lam et al. [14]The fluctuation of V^mcaat 1.5 MAC may partly be ascribed to the nociceptive stimulation. The change of flow velocity induced by surgical stimulus has been recently reported by von Knobelsdorff et al., [15]showing that flow velocity significantly increases, lasting for 3 or 11 min after skin incision in 1 and 2 MAC isoflurane-anesthetized patients. These changes were possibly a result of the increased cerebral metabolism (i.e., arousal response) but not to the increase of MABP. [15]Bisonnette et al. [21]showed that V^mcaremained stable during 1.0 MAC isoflurane anesthesia, with concomitant use of epidural local anesthetic. The fluctuation of V^mcaat 1.5 MAC may also partly be ascribed to changes in CBF associated with MABP changes because of the possibility of impaired autoregulation. [29]Taken together, the changes of V^mcaobserved appear to be the net results of the vascular and metabolic action of anesthetics and the effect of nociceptive stimulation. Our previous observation [11]that CBF equivalent was preserved is in agreement with the present one in view of existence of flow-metabolism coupling during 1.5 MAC volatile anesthesia.

In summary, the current study shows that no decay in CBF over time occurs in the patients during surgery under prolonged (3 h) inhalation of volatile anesthetics (1.5 MAC). The fluctuation of CBF observed during burst suppression in EEG at 1.5 MAC during isoflurane and sevoflurane anesthesia indicates the existence of flow-metabolism coupling.